I represent the old ORNL tradition about nuclear technology. Oak Ridge scientists quickly rejected the idea of a sodium cooled reactor (1947 to 1950). Indeed Eugene Wigner who was the first post War research director at ORNL and who held the original patent on the Liquid Sodium Cooled Reactor, did not like his own invention. The original Oak Ridge air craft reactor was sodium cooled, but it was developed at K-25 rather than ORNL and the K-25 engineers who were developing the K-25 reactor project, quickly realized that the sodium cooled air craft reactor design had deep safety flaws including a positive coefficient of reactivity, and of course the insidious dangers of a liquid sodium coolant. Ed Bettis and his associates quickly bailed out on the liquid sodium cooled reactor design, and developed the Molten Salt Reactor concept. it was not by accident that the original K-25 MSR concept featured liquid fluoride salts, If K-25 chemists and engineers knew anything, they knew fluoride chemistry. Fluoride salts are, of course, quite safe compared to liquid sodium. The original MSR concept featured a negative coefficient of reactivity. Indeed if American Reactor development policy had been guided by strictly rational considerations from 1950 onward there would have been no more money spent on liquid sodium cooled reactors. As it is over 20 billion 2009 dollars has been spent on Sodium cooled fast breeder research without the development of a single American commercial Fast Breeder prototype.
By the early 1950's my father, C.J. Barton, Sr., was exploring the chemistry for a Liquid Chloride Fast Breeder that would have been safer, and far more practical than any LMFBR design concept. Had the LCFR rather than the LMFBR been chosen as the major direction for United States Breeder fast breeder research, and been supported at the level that the Liquid Sodium Fast Breeder was to receive, I have little doubt that an American Fast Breeder Reactor would have been developed during the 1960's. Unfortunately that was not the case, and instead over 20 billion 2009 US dollars were tossed by the United States Government down a rat hole marked Liquid Sodium Fast Breeder Reactors.
In addition to ORNL studies of the LCFR concept, a 1974 British study reached favorable conclusions about the LCFR concept, and a 1978 Swiss report indicated that the LCFB was a promising design.
In his 1982 UCLA dissertation by E. H. Ottewitte, who had participated in the Swiss study stated:
Molten salts compete favorably with liquid metals: they exhibit thermal conductivities intermediate to water and the poorer of the liquid-metal. Their specific heat capacities parallel water's. Furthermore, an intermediate coolant of molten salt should more closely match the primary salt in physical proper-Ottewitte concluded
ties, thereby reducing freezing and thermal stress problems. They will cost far less then liquid metals.
Fast molten chloride reactors have been cursorily considered before but mainly for the U/Pu fuel cycle. The ORNL MSR program showed the feasibility of fuel salt circulation. The combination of that experience and MCFR research (out-of-pile experiments and theoretical studies, so far) provide a basis for believing theIn 1992 Ottewitte listed some of the advantages of the LCFR:
concept will work.
Chemical stability and corrosion of molten salts are fairly predictable. Low vapor pressure of the salts enhances safety and permits low pressure structural components.
Molten fuel state and cooling out-of-core simplify component design in a radiation environment. They forego complicated refueling mechanism, close tolerances associated with solid fuel, and mechanical control devices. Molten state and low vapor pressure of the salts also offer inherent safety advantages.
Some salient advantages of the MCFR concept are :The American energy establishment, never looked seriously at the LCFR as a viable alternative to the liquid sodium fast breeder. If it had there is a likelihood that the LCFR would fave compared favorably to all LMFBR concepts including the IFR. The question would have been if a fast breeder reactor was needed at all. The LFTR offers sustainable breeding with the advantage of operating in the thermal/epithermal neutron range. The availability of a slow neutron breeding process potentially offers an enormous scalability advantage. A limited nuclear fuel supply would pose a far more serious challenge to the scalability of the IFR or a LCFR than it would to a two fluid LFTR with or without graphite.
1. Simplicity : no control rods, fuel handling mechanisms, fuel elements or associated structures . Very uncluttered: should maximize test space and facilitate access thereto . Fluid fuel can be transferred remotely by pumping through pipes connecting storage and reactor .
2. MSRs don't refuel or reprocess, just add fuel and process out wastes . Continuous
processing and refueling would minimize reactor downtime . Can usefully consume all fuel forms, simplifying fuel supply while simultaneously solving other people's
problems.
3. MSR is the safest concept of all due to very strong negative temperature coefficient. No gaseous hydrogen can possibly evolve from fuel or primary coolant . Fuel already molten and handled by system . Simple design technique makes boiling impossible. Continuous removal of fission products reduces their heat source by two orders of magnitude: consequently, natural circulation suffices for emergency cooling, thereby greatly reducing the designated evacuation area . Also, under any off-normal conditions, the liquid fuel can be channeled to a continuously cooled drain tank, in a short time.
4. Very fast neutron spectrum in an annular core engenders high neutron fluxes, driving inner and outer thermal neutron flux traps, each variable in size and neutron energy spectrum by means of molten salt composition. Elimination of fuel cladding and structural material significantly improves the neutron economy of the reactor: more neutrons are available for applications.
5 .Elimination of pressurized and pressure-evolving components inside the containment, reducing risk of containment failure.
IFR concepts during the 1990's included the notion of factory constructed modular reactors. But no assessment has been yet offered on possible interaction between IFR size and safety. Safety is a far more critical issue for the IFR than for a MSR for obvious reasons. The NRC has not yet developed safety concepts for either the LFTR or the IFR, but given the implications of major safety problems such as a coolant leak or a core rupture, IFR safety requirements are likely to be far more stringent, and costly to meet. NRC safety requirements will probably include the sort of massive safety related site construction requirements that favor economies of scale.
Compared to both the LFTR and its chloride salt cousin the LCFR, the IFR would pose significant safety hazards, and would be larger and probably more expensive to build in a factory setting. It would also pose significantly more control issues, and a system of defense in depth against sodium related safety concerns is likely to add to reactor complexity and cost. The IFR would require over 10 times as much fissionable materials as a graphite moderated LFTR making the LFTR a far more attractive candidate for the role as a mass produced, widely deployed global warming fighting technology. Indeed, factor produced small LFTRs may offer significant cost advantages over all other post-carbon energy technologies.
Politicians and scientists often feed each other's insecurities. Since the 1972 firing of Alvin Weinberg for political reasons that included his MSR advocacy and criticism of the safety of LMFBRs the American scientific community has been cowed by the orthodox dogmas of the US Department of Energy. The US government did supply limited support to the concept of a thorium fuel cycle thermal molten salt breeder from the 1950's to 1970's, and although that reactor concept received no more than 4% of the money wasted is such a profligate fashion by on the liquid sodium breeder reactor concept, that project proceed in an outstandingly successful fashion, and would have become the crowning success of the United States nuclear program if Washington had been willing to back it with even 25% of the money it wasted on the LMFBR concept.
Posts
14 comments:
Charles, an interesting analysis.
The IFR concept includes metal fuels (a ternary alloy of U-Pu-Zr) which provide a -ve coefficient of reactivity. Of the other advantages you list for the MCFR, the IFR also provides sufficient cooling via natural circulation of the sodium, and of course the sodium will not boil (criticality will be lost well before such temperatures are reached). The IFR operates at near-atmospheric pressure.
There are lingering concerns about the safety of sodium and the possible dangers of sodium leaks -- something I will address in a post on my site in the coming week or two. It is an issue worth examining in more detail than I have room to do here, but sufficient to say this is a card far too overplayed.
The LCFB is an interesting concept and I wish it had been developed more fully 30 years ago. But the fact is that most funds were directed to the LMFBRs and the result was ultimately the exemplar 10 year project at Argonne from 1984 to 1994 that developed the Integral Fast Reactor concept and mostly proved it up (the last step remaining is commercial scale pyroprocessing, which has seen ongoing development).
I know it has left a bitter taste in your mouth to think of the unnecessary lost decades that followed the abandonment of serious LFTR research at ORNL. I, like you, wish it had been otherwise, and that we now had a commercialised LFTR and even LCFB. Realities will change in the coming decades, but for now, we have the S-PRISM, an IFR-based design, ready for build. It represents an excellent concept, as readers of Nuclear Green will appreciate if they take the time to read over my blog (http://bravenewclimate.com) and of course read Tom Blees book, Prescription for the Planet.
Whether the LFTR or IFR (or LCFB) is the superior technology is largely an academic matter. The bottom line is that all are excellent designs and represent, to me and many others, the future of nuclear energy, the most realistic way to solve the climate and energy crises, and a source of great energy prosperity for many future generations.
All should therefore be pursued vigorously, and I make it a point not to try to convince anyone that one design trumps another. We cannot really know that until large-scale build-out is well underway. My hunch is that both will play significant roles.
If we are stuck with the situation where one must choose either the IFR or the LFTR when scrapping for funds, then I think we, as a community of nuclear environmentalists, have a problem. That will mean that insufficient attention is being directed towards nuclear power, and that the possibility of a worldwide supply of carbon-free energy generation has been squandered. At that point, climate will damn all losers.
Berry, Onje of my complaints about renewable supporters is that they ignore potential debates among themselves. I think it is better for use to lay the cards out on the table, and let the public know that there may be choices to make. Part of the bitterness at ORNL stemmed from the fact that hundreds of millions of dollars that could have been spent or MSR development, was spent instead on the poorly designed Clinch River Breeder Reactor that was built in Oar Ridge, but was not an Oak Ridge project. There was also the feeling that Argonne and INL stabbed ORNL in the back as part of the IFR deal.
Being a huge fan (as you know, Charles) of chlorides fast systems, I' m very glad you wrote a post about.
Actually, I don't see a real competion between IFR and LCFR. I think the first application of LCFR will not be to harness the uranium/plutonium cycle, but rather to burn transuranics LWR wastes and produce the first start-up of uranium 233 in the thorium blanket for the (thermal or epithermal) fluorides reactors (as Kirk Sorensen envisages), a goal very hard to achieve with IFR; if necessary, only later the system can switch completely to a "pure" U-Pu (or TRU) cycle, there is no need to do that today
I wonder if a same fast chloride core can be easily "converted" - with marginal modifications - to switch from a thorium to a depleted uranium blanket working in the U-Pu cycle
Unfortunately Alex, I think it would be rather simple to change an LCFR to a U/Pu breeder, and a very good one. That's one of the reasons that despite my strong interest in LCFR as a start-charge breeder for LFTR, I think the deployment of LCFRs should be very limited and tightly controlled. These are not reactors we want in mass deployment worldwide.
To be more specific, Kirk, are you concerned about the possibility in a same LCFR core to change the thorium blanket with a DU one (as far I understand you claim it's very easy to do that or am I wrong?) or in general about LCFR tech even with a simply thorium blanket ?
[Anyway, it's my opinion that LCFRs working with the U/Pu cycle, expecially in enriched chlorine versions, is even more proliferative than sodium fast reactors...]
An other question Kirk, have you got a rough idea about the costs to build a couple of MSR prototypes in the range of 100 MW thermal of power (in order to test the start up technologies), for example a reactor emulating '60 MSRE design, or an epithermal moderator free version and a chloride fast reactor? I guess it's someting in the range of 15,000 $ per kW electric....any thoughts about?
Just on the ICFR...I've been off/again on/again thinking about this as well every time we have a proliferation discussion.
I think it's clear that any US LCFR will be tightly controlled, maybe even Federally owned (DOE) and operated. But Kirk's idea is a good one: LCFRs specifically integrated as into the LFTR paradigm for start up charge. We many only need them for 10 or 20 years then as strictly LWR SNF burners.
Sorry, a little off topic.
I urge everyone to visit Barry's blog and chime in. It's an excellent one and, provides a lot of usefull data.
I do not thing that IFRs are necessarily counter-poised. R&D should be expanded, not shifted, from one reactor paradigm to the the other.
I'd like to see more discussion on sodium fire/safety issues.
David
Charles, I don't know the history of the Argonne/ORNL interaction in detail, but that's in the past, for good or bad. The IFR is now close to commercialisation via the GEH S-PRISM. It would be crazy not to pursue this. The BN-800 is being built in Russia, and a couple are probable for China, all going well, so the LMFBRs aren't going away.
A key feature of the IFR is pyroprocessing, which makes it inherently more proliferation resistant than the LCFB. Sure, you can run an IFR on a short cycle and build an expensive and heavily shielded PUREX-type plant to get weapons-grade material from an IFR, but there would be only one reason to do this -- so intent would be patent.
I reiterate, I support both the IFR and LFTR. I don't see any compelling reason to dismiss one tech and accept the other just because of the philosophical need for a debate on relative merits. I just don't think these are provable until large-scale deployment anyway. If you are worried that IFR deployment will be used as a case to ignore LFTR RD&D, then I think it is up to the nuclear community to emphasise to decision-makers the benefits of a diverse sustainable nuclear mix. The very fact that one uses DU and the other Th is perhaps reason enough to push for both.
Barry, I have Posted a responce to some of Tom's comments on your blog. While my post does not answer all of Toms points or yours for that matter, I do indicate that we need a whole lot more information plus the posting of Data via peer reviewed studies before you would have a strong case.
" A key feature of the IFR is pyroprocessing, which makes it inherently more proliferation resistant than the LCFB "
Frankly, I don't understand why the IFR should be more proliferation resistant than LCFR, expecially working with a nice thorium blanket
I want make clear I'm not against IFR, but I'd guess the LCFR needs less infrastructures and is less complex than the IFR, so indeed easier to build and operate. We should start with R&D building a couple of prototypes of both (in the range of ~ 100 MW thermal, at max) and only later - I believe - try to understand which of two systems is the best one, in terms of safety, proliferation resistance, complexity and of course economics, i.e. costs of building and operation. I'd think any MSR system is quite superior than IFR, but I want to see operated "real" (and not "paper") reactors
Alex P -- pyroprocessing cannot separate Pu from the minor actinides. IFR consists of a SFR with metal fuels and and on-site pyroprocessing. If you wish to proliferate with an IFR, you'd have to build a highly specialised, heavily-shielded off-site PUREX-type facility. There is no reason to do this unless you wish to make pure Pu for bombs. That's why (in brief).
Charles, I'll provide you a bunch shortly.
I don't discuss the potential benefits of IFR proliferation resistance - even if, unfortunatly, with any nuclear system when you handle nuclear spent fuel there is always even a remote chance to proliferate - but with MSRs we can achieve even better anti proliferation barriers, in particurally with the thorium cycle; we'll never need (the reactor doesn' t need to refuel nor to reprocess) to extract the fuel from the reactor and handle it in an external (even on site) reprocessing plant
Stepping back a second, "proliferation". I hate the discussion but it's politically necessary.
Neither system is going to be a plutonium factory for WMD. Or they will be (if you run the LFTR/LCTR on DU for example and not T).
Ultimately the paradigm of the very easy to build, cheap and small graphite brick reactor is going to be the main source of Pu-239. Even if I HAD a IFR or LCTR...it would be a total waste to use these as WMD.
Secondly, and far more importantly...so what? Really? We build a security Gen IV reactor FOR WMD for the DofD again SO WHAT? The U.S., unfortunately, spends money on this openly and for maintaining the US huge store of nuclear WMD. This will NOT change one iota so again, WHO CARES?
Ah, but you say overseas we can't be so sure? See: research reactor. The cat is out of the bag. A *political/diplomatic* and, I assume, other methods, are the only way to stop proliferation. It's one of policy, not technology anymore. And it will always be.
Look, the technology cannot be restricted. IFR and LFTR papers are everywhere so ANY one can, in theory, assemble the nuclear, power, and material engineers to build one. Use the research reactor and enrichment facilities to make start up charges, and you are off and running.
WMD should have NO brakes in building a IFR or LFTR in the U.S. or Australia. None. Zero. Zip. Any such facility, the R&D engineering labs where they will be first developed and then deployed, will have to have excellent security. That is all we have to worry about.
David
David you have got it. The whole proliferation argument is a scam. North Koreas has all of the proliferation tools it needs, a cold war era gas cooled graphite pile which if skillfully managed can reliably yield weapons grade plutonium. Fortunately the North Koreans don't know how to do the skillful management yet.
Back again to the LCFR technology : just curious, given the fact that sodium for solid fuel fast reactors was the principal choice mainly for corrosion issues, is there any corrosion problem for chlorides salts in LCFR?
Post a Comment